Well, I'm working now on creating memory dump. and send it forward for more knowledgeable people, however, as you might notice, different people, with different hardware, and even different version 7.0 and 7.1, have the same problem.
Even if lighttpd / php / some script / whatever misbehaves, system should not be halted by such userland proccess. Regards, Yury. On Mon, Sep 22, 2008 at 6:33 PM, Nash Nipples <[EMAIL PROTECTED]>wrote: > >> It looks like the freebsd-sendfile is broken. I had the same problems > >> the last days and now I know the source of the problem. Have a look on > >> this PR: > >> > >> http://www.freebsd.org/cgi/query-pr.cgi?pr=125592 > > >Yeah, I looked into this, but I do not see why this has anything to do > >with > >sendfile in FreeBSD. > >When you upload a file, you (the server) don't *send* anything, you > >*read*, so > >sendfile(2) shouldn't even come into play. Why this magically starts > >working > >when changing the write backend, is a mystery, but most likely the answer > >lies in PHP. > > >If you can, use a different upload script for testing, one that is not > >done > >via PHP, but plain CGI. There should be some available on the net. > >-- > >Mel > > im in a serious doubt that a write() family function could have possibly > stop something from responding when login is still accepting read() buffer > > please dont tell me that a userland process can possibly halt everything > unless there is a serious harware malfunction. > > i think this particular case needs solid debugging and investigation. > > nash > > > > _______________________________________________ > freebsd-questions@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-questions > To unsubscribe, send any mail to " > [EMAIL PROTECTED]" > _______________________________________________ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "[EMAIL PROTECTED]"
